The following tests were run with an NVIDIA Tesla K40c GPU on a system with an Intel Core i7 920 2.67GHz CPU.

Source code for the benchmarks can be found under the perf directory. All benchmarks were compiled with optimizations enabled (i.e. "gcc -O3").

Accumulate

Count

Inner Product

Merge

Partial Sum

Partition

Reverse

Rotate

Set Difference

Sort

Transform

Unique